Correspondence: Professor T. F. Baskett, Department of Obstetrics and Gynaecology, 5980 University Avenue, Halifax, Nova Scotia B3J 3G9, Canada.
The origin of Naegele's rule to calculate the expected date of confinement is reviewed. It is possible that the rule has been misinterpreted, resulting in an earlier estimated date of delivery with implications for induction of labour for post dates pregnancy.
The estimation of gestational age and the expected date of delivery is central to the social and clinical essence of pregnancy. Generations of midwives and doctors have learned Naegele's rule, based on menstrual dates, to calculate gestational age. Recently, with the widespread use of ultrasound in many countries, calculations using menstrual dates have come to be regarded as less accurate than those based on ultrasound measurement, although this is discounted in a recent review1. Thus, it remains clinically relevant to continue to use Naegele's rule, or some variation thereof, particularly in the vast majority of pregnancies that occur worldwide without the presumed benefits of ultrasound. This article reviews the origins and evolution of Naegele's rule, its possible misinterpretation in modern obstetrics, and the clinical implications of its erroneous application.
Naegele's rule assumes a 28-day-cycle with ovulation on day fourteen. By adding seven to the first day of the last menstrual period and counting back three months the expected date of confinement is reached. This is the method by which Naegele's rule has been calculated during this century. However, it was not always thus and Naegele did not invent this rule or lay claim to have done so.
For many centuries it has been accepted that the normal gestation period for humans is ten lunar, or nine calendar months. This was reinforced by the New Testament account of the birth of Christ and the nine month gestation from the Feast of the Annunciation in March until Christmas Day. It was Hermann Boerhaave (1668–1738), Professor of Botany and Medicine at Leyden University, who first set down the calculation from which Naegele's rule evolved. In the Academic Lectures of Hermann Boerhaave, which was edited and annotated by Albert Haller and published in 1744, the relevant passage from his lecture On Conception translates as follows2:
‘Women for the most part are impregnated after the end of their period: Numerous experiments undertaken in France confirm this: for of one hundred births altogether, ninety-nine came about in the ninth month after the last menstruation by counting one week after the last period and by reckoning the nine months of gestation from that time. For, at that time the uterus is purged and empty, and the plethora are drained out…’
It should be noted that Boerhaave said ‘…counting one week after the last period…’From this statement we cannot be sure whether he meant one week after the start, or one week after the end of the last period.
Franz Carl Naegele (1778–1851), Professor of Obstetrics at the University of Heidelberg, quoted this section in his 1812 text (Fig. 1) and also made his own observations on the ability and timing of conception in women3: ‘According to experience, the woman in her reproductive phase does not always have the same capacity to conceive. The time the woman is most likely to conceive is immediately after menstruation’. At this point in his text, Naegele cites and quotes Boerhaave: ‘Women, says Boerhaave, always conceive after the last menstruation and scarcely at any other time’. Naegele then continues: ‘The usual calculation of the duration of pregnancy, namely, starting from the last menstruation is correct in most instances; and conception within the last third of the cycle or in the second half between two periods is unusual and an exception to the rule.‘3 Thus, it is clear that Naegele attributed this method of calculation to Boerhaave. It can also be seen that the wording lacks precision, so that one could interpret conception as occurring either seven days after the start or seven days after the end of the last period—although the latter may be plausible as it is mid cycle and therefore likely to coincide with ovulation. Speert4 believes that it was Gunning Bedford who first attributed the rule to Naegele. Bedford, the Professor of Obstetrics and Diseases of Women and Children in the University of New York, did so in his 1872 text The Principles and Practice of Obstetrics5:
‘I have, for several years, adopted a rule, which, I believe, was originally suggested by the celebrated Naegele; with some exceptions, I have found it generally quite reliable, and far more satisfactory in its result than any plan which has yet been proposed. Imagine, for example, the termination of the last menstrual period to be on the tenth day of January; then count back three months, which will correspond with the 10th day of October; now from the 10th of October add seven days—this will bring you to the 17th day of October— the day on which labour will commence. This, I repeat, has, according to my observation, proved a most satisfactory test; and I, therefore, commend to you with much confidence. According to this mode of computation, the short and long months are taken promiscuously together, and the addition of seven days constitutes the average difference in time.’
In this interpretation of Naegele's rule Bedford has taken the date to which the seven days should be added as the end of the period. Neither Boerhaave or Naegele were this specific in their writings. One American text of the mid 19th century retained the vague advice of calculating ‘from the last menstrual period’6. The influential Williams Potts Dewees of Philadelphia recommended that the calculation should be from the ‘cessation of the menses’ and stated that ‘it is always best to allow a little latitude, beyond the time which the mere stopping of the menses would indicate…‘7. By the late 19th and early 20th century the standard American texts advocated Naegele's rule, which they interpreted as adding seven days to the start of the last menstrual period8,9. However, a British text from 1921 advises adding five days to the last day of the last menstrual period and counting back three months10.
The calculation of gestational age and the expected date of delivery is not just a social issue but has considerable medical implications associated with induction of labour. It is common obstetric practice to induce labour 7–14 days past the expected date of delivery to obviate the very rare risk of fetal death, presumably due to failing placental function. Induction rates as high as 16% are recorded for this indication11. Gestational age based on dates assumes an invariable 28-day-cycle and that ovulation always occurs mid-cycle. While this is a reasonable working assumption there are many variations between and within women. Estimates of gestational age based on basal body temperature to detect ovulation and presumed conception, compared with menstrual dates reduces the number of pregnancies that go past term12. In a large population-based study from the Swedish birth registry involving 427,582 singleton pregnancies the mean, median and modal durations of pregnancies were 281, 282 and 283 days respectively13. The modal duration was felt to be the most accurate as it reduced the influence of pathological pregnancies at the extremes of distribution. This group suggested that most pregnancies undergoing post term induction are not in fact post term when assessed by ultrasound13. Mongelli et al.14 found ultrasound to be superior to ‘certain’ menstrual dates in prediction of the actual date of delivery in 34,249 singleton pregnancies. Hutcheon and Kearney15 observed a lack of concordance among clinicians using the same menstrual and ultrasound dates and found a computer system using the same data to be more accurate.
As Olsen has pointed out, ultrasound estimation of gestational age postpones the expected date of delivery compared with that calculated from menstrual dates and therefore reduces the number of post term pregnancies in which induction of labour is considered16. He claims that if three days were added to Naegele's rule, this might achieve the same result. Similarly, Backe and Nakling found that in women with regular and known cycles, Naegele's rule predicted the expected date of delivery three days too early17. Thus, from ultrasound, basal body temperature, and the modal duration of pregnancy from the Swedish birth registry, there is much to be said for refining Naegele's rule and adding ten days to the first day of the last menstrual period rather than seven. It is possible that Boerhaave and Naegele have been misinterpreted, and their original rule may have been to add seven days to the end rather than the beginning of the last menstrual period, which, with respect to induction of labour for post dates pregnancy, would achieve the same result as ultrasound.
We are grateful to Mr R. Butler, Toronto for the translation of relevant passages from the Academic Lectures of Hermann Boerhaave and to Miss P. A. Want, Librarian, RCOG, for assistance in retrieval of the works of Boerhaave and Naegele.